Harvey LAB-AA - Intellectual Property: leaderboard
Metric: Criterion Pass Rate (%). Source: artificialanalysis.ai. 43 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Muse Spark 1.3 (xHigh) | 96.82 |
| 2 | Claude Opus 5 (Adaptive Reasoning, Max Effort) | 96.82 |
| 3 | Grok 4.5 (High) | 96.47 |
| 4 | Kimi K3 (Max) | 95.76 |
| 5 | Claude Opus 4.8 (Adaptive Reasoning, Max Effort) | 95.76 |
| 6 | GLM-5.2 (Max) | 95.05 |
| 7 | Claude Sonnet 5 (Adaptive Reasoning, Max Effort) | 95.05 |
| 8 | Gemini 3.7 Flash (High) | 94.7 |
| 9 | MiniMax-M3 | 93.99 |
| 10 | GPT-5.5 (xHigh) | 93.99 |
| 11 | GPT-5.6 Luna (Max) | 93.64 |
| 12 | Gemini 3.6 Flash (High) | 93.29 |
| 13 | Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) | 92.58 |
| 14 | GPT-5.6 Sol (Max) | 92.23 |
| 15 | GPT-5.6 Terra (Max) | 90.81 |
Interactive version: theaggregate.ai/benchmark?slug=harvey-lab-aa-intellectual-property · How It Works · Data refreshed daily, snapshot 2026-09-19.